Introduction

Row

The COVID-19 pandemic has given rise to speculations regarding the prices of property in Melbourne. Many have speculated about a Melbourne “housing price bubble” that will soon burst, causing a decline in housing prices. Others have speculated that prices for Australian property will continue to grow. The motivation of this report is to discuss in-depth the level of price volatility of property in Melbourne with attention to many variables such as location, property attributes and time.

Missing data

Missing data in Regions

Missing data in council areas

Question 1

What is the relationship between property distance from the CBD and price?

Row

Distance-price relationship

Distance-price relationship faceted by region

Number of properties sold in each region

Average distance-price relationshiop

Interactive plot

Question 2

Which regions and council areas have the most expensive housing in Melbourne?

Row

Average price across regions over time

Spread in property prices

Proportion of average property prices

Proportion of house price according to council area

Top 5 most expensive Council Areas

Grouped by type of property

Price-landsize-rooms relationship

Interactive

Question 3

What is the regression model used to predict the value of property?

Row

Price relationship with variables Rooms, Bathroom, Landsize

# A tibble: 4 × 5
  term         estimate std.error statistic   p.value
  <chr>           <dbl>     <dbl>     <dbl>     <dbl>
1 (Intercept) -64493.    11914.       -5.41 6.26e-  8
2 Rooms       262913.     4857.       54.1  0        
3 Bathroom    251011.     6663.       37.7  2.03e-301
4 Landsize         2.95      1.18      2.51 1.20e-  2

Single linear regression model Price ~ Room

Single linear regression model Price ~ Bathroom

Single linear regression model Price ~ Landsize

Question 4

How do the key variables contribute differently to the price of the property in different suburbs?

Row

Regression

# A tibble: 7 × 5
  term         estimate std.error statistic     p.value
  <chr>           <dbl>     <dbl>     <dbl>       <dbl>
1 (Intercept) 3977541.   844301.      4.71   0.00000367
2 Rooms        245567.    76404.      3.21   0.00144   
3 Bedroom2      74071.    79013.      0.937  0.349     
4 Bathroom      73737.    34398.      2.14   0.0328    
5 Landsize         13.9      27.0     0.516  0.606     
6 YearBuilt     -2024.      429.     -4.72   0.00000353
7 Distance         NA        NA      NA     NA         
# A tibble: 8 × 5
  term          estimate  std.error statistic       p.value
  <chr>            <dbl>      <dbl>     <dbl>         <dbl>
1 (Intercept) 9544036.   1774648.       5.38  0.000000145  
2 Rooms        433736.    109717.       3.95  0.0000949    
3 Bedroom2      63477.    110292.       0.576 0.565        
4 Bathroom     291676.     50889.       5.73  0.0000000229 
5 Landsize         -2.52       9.17    -0.275 0.784        
6 YearBuilt     -5294.       879.      -6.02  0.00000000469
7 Distance     132282.     86415.       1.53  0.127        
8 Car           46570.     41093.       1.13  0.258        

Final regressions

# A tibble: 4 × 5
  term        estimate std.error statistic  p.value
  <chr>          <dbl>     <dbl>     <dbl>    <dbl>
1 (Intercept) 4020637.   834932.      4.82 2.25e- 6
2 Rooms        315413.    19799.     15.9  1.25e-42
3 Bathroom      81373.    33500.      2.43 1.57e- 2
4 YearBuilt     -2044.      424.     -4.82 2.19e- 6
# A tibble: 4 × 5
  term        estimate std.error statistic  p.value
  <chr>          <dbl>     <dbl>     <dbl>    <dbl>
1 (Intercept) 9754083.  1669210.      5.84 1.24e- 8
2 Rooms        500103.    37537.     13.3  1.30e-32
3 Bathroom     310430.    49352.      6.29 1.02e- 9
4 YearBuilt     -5181.      838.     -6.18 1.93e- 9

Linear regression for Brunswick

Linear regressions for South Yarra

Residuals

Regression equations

\[\widehat{\text{Price}} = 4020637 + 315413~\text{Rooms}+ 81373~\text{Bathroom}- 2044~\text{YearBuilt}\]

\[\widehat{\text{Price}} = 9754082 + 500103~\text{Rooms}+ 310429~\text{Bathroom}- 5180~\text{YearBuilt}\]

Question 5

Which suburb and region provided the most houses for different classes of families?, with classes being: Lower class, Middle Class and Upper Class

Row

Lower class region

[1] 188

Lower class suburb

Middle class region

[1] 12774

Middle class suburb

Upper class region

[1] 8859

Upper class suburb

Conclusion

Row

  • Property price tends to increase with a decrease from the distance from the Melbourne CBD. However, other location variables such as council area and region seemed to have a larger impact on property pricing.

  • The most expensive region, in terms of housing, is the Southern Metropolitan while Boroondara, which is in that same region, is the most expensive council area.

  • We found how the number of rooms, bathrooms, and Year built have different impacts on the price of house in Melbourne and particularly how they have different impacts on price in Brunswick and South Yarra

  • We concluded that the region and suburb that sold the most houses for lower-class families were Western Metropolitan and Footscray. For middle-class families, they were Northern Metropolitan and Reservior. Finally, for upper-class families, they were Southern Metropolitan and Glen Iris.